Obesity is a complex multifactorial disease, characterised by excess accumulation of body fat [1, 2]. Since 1980, prevalence of obesity has nearly doubled [3]. This major increase in prevalence of obesity is particularly concerning as fat mass gain can promote other co-morbidities such as type 2 diabetes and certain forms of cancer [1, 2].
Both genes and environment substantially contribute to changes in body weight across the life span [4, 5]. Landmark studies on twins and non-related individuals showed that genetic differences can account for 50–70% of the disparity in body mass among the population [6–8]. On the other hand, our genetic makeup has remained relatively unchanged over the past century [4], and yet, rates of obesity have been constantly increasing [3]. Consequently, several researchers suggested that urbanisation and modernisation have both contributed to the reduction of physical activity and facilitated availability of energy-dense foods [5], which collectively led to the creation of an obesogenic environment [9]. While both genes and environment seem to influence body mass [5, 8], gains in body fatness are ultimately caused by prolonged periods of positive energy balance, a state where energy intake exceeds energy expenditure [10]. Conversely, a reduction in body fat can be induced through a state of negative energy balance over prolonged periods [11]. Notably, achieving a clinically meaningful 5–10% reduction in body mass showed positive effects on several health markers such as blood pressure, blood lipids and blood glucose [12]. On this note, a state of negative energy balance can be induced via exercise, dietary restriction or by a combination of both [10]. When combined, exercise and dietary restriction seem to synergistically enhance body mass loss [13]. By contrasts, interventions using exercise alone demonstrated a high degree of intra-individual variability and compensatory changes, with only modest reductions in body mass that are often below-predicted [14–16]. For example, Borer [17] found that in absence of dietary restriction, daily exercise-induced energy expenditure (ExEE) of 400kcal yielded only 30% of the theoretically predicted body mass loss over the course of a month. Similarly, Barwell [15] found that 7 weeks of ExEE resulted in highly divergent changes in body mass ranging between − 5.3kg to + 2.1kg in sedentary women. Initially, the below-predicted reductions in body mass were mainly attributed to poor adherence to exercise interventions. However, Church and colleagues [18] compared actual to predicted body mass loss across three doses of ExEE (4, 8 and 12 kcal/kg/week) during 6 months of supervised exercise. Reductions in body mass nearly matched the predicted body mass losses in the 4 and 8 kcal/kg/week groups, however, in the 12 kcal/kg/week group, the actual body mass loss was significantly lower than the theoretically predicted. Interestingly, greater reductions in body mass were observed in the 8 kcal/kg/week group than the 12 kcal/kg/week group, despite a lower total ExEE [18]. These findings indicate that ExEE may modulate appetite, potentially promoting a compensatory response by increasing energy intake (EI).
Eating behaviour in response to exercise seems susceptible to substantial intra-individual variability [19–21]. After 7–14 days of structured exercise, energy intake starts to track the disruption in energy balance and compensates for ~ 30% of the ExEE [22, 23]. Some individuals seem to be more susceptible to compensation via EI than others [24, 16]. Nevertheless, this becomes paramount when considering that in response to a perturbation in energy balance, EI can vary by up to 100% within the same day [25].
Notably, components of energy balance are often assumed to be static and hence, an increase in ExEE is expected to result in a linear increase in total daily energy expenditure (TDEE) [10]. However, energy balance is dynamic, and when one component of TDEE is perturbed, numerous biological and behavioural factors can be affected resulting in partial compensation in other components of TDEE such as resting metabolic rate (RMR), thermic effect of food (TEF), exercise-induced energy expenditure (ExEE) and physical activity energy expenditure (PAEE) [26].
In sedentary individuals, RMR accounts for 60–70% of TDEE and can be defined as the amount of energy required for homeostatic processes [27], whereas TEF is the energy expenditure related to digestion and storage of food accounting for 8–15% of TDEE [11]. Energy expenditure yielded from physical activity is the third main component of TDEE and can be furtherly divided into ExEE and PAEE [10]. The former is the energy expenditure yielded from volitional exercise, whereas the latter is energy expenditure related to non-exercise activities such as maintenance of body posture, ambulation, fidgeting and other spontaneous activities [28, 29]. In individuals who regularly take part in purposeful exercise, ExEE can account for maximally 15–30% of TDEE, whereas on a population level, ExEE is believed to be negligible [29]. By contrast, PAEE can account for 6–10% of TDEE in individuals with sedentary lifestyles and up to 50% in highly active ones [28]. Interestingly, mainly depending upon occupational activity, PAEE can vary by as much as 2000kcal/day between individuals with similar body size, body composition, sex and age [30, 27].
When a state of negative energy balance is elicited via dietary restriction, a reduction in PAEE is often observed [28, 31]. Conversely, exercise-induced compensations in PAEE have been poorly explored16,19,20. Long-term studies (12 weeks to 16 months) show little evidence in support of an exercise-induced compensation in PAEE [32–35]. By contrast, medium-term (16 days to 12 weeks) and short-term (up to 16 days) trials provide more divergent findings [35–40]. For example, Manthou and colleagues [40] found that 8-weeks of ExEE resulted in divergent body mass changes ranging between − 3.2 and + 2.6kg in overweight women. The researchers attributed the failure of exercise to induce body mass loss to the exercise-mediated reduction in PAEE. In a similar study involving overweight women, Colley and colleagues [41] observed a 22% decrease in levels of PAEE from baseline following 8-weeks of ExEE of 1434 ± 237Kcal per week. In a randomized control trial involving lean and overweight women, Schutz and his colleagues [36] reported decreased levels of PAEE only in overweight women after 4 weeks of low-intensity ExEE. Short-term studies involving overweight men and lean women showed no exercise-mediated compensation in PAEE [35–42]. On the other hand, a degree of compensation in PAEE was observed when lean men performed multiple bouts of exercise per day at moderate intensity [39, 43].
Exercise studies rarely enrol women [44, 45]. Therefore, similar studies that explore the impact of exercise on levels of PAEE in overweight and obese women are lacking [46]. In fact, women seem to experience a proportionally lesser reduction in body mass than men [47]. The hypothesis for this study was that an acute exercise-induced perturbation in energy balance would decrease PAEE, and that this compensatory response would be larger in women than in men.